Predictors of Injurious Falls and Fear of Falling Differ

Abstract

Objective: The objective of the study is to identify the intrinsic, psychosocial and lifestyle factors, which, over time, predict the incidence of having a fall requiring medical attention (injurious fall) or of acquiring a fear of falling (FOF). Method: Data from 1,000 participants in the Melbourne Longitudinal Studies on Healthy Ageing (MELSHA, 1994-2005) were analyzed using cox regressions and hazard ratios. Results: The predictors of injurious falls (n = 900, events = 200) were increasing age, slower gait speed, and being depressed. Main predictors of developing a FOF (n = 855, events =117) were increasing age, cognitive impairment, reduced social activity, and gender. A history of falls at baseline did not predict acquiring a FOF nor did FOF predict a future fall. Discussion: The profile of the person who will have an injurious fall differs from the profile of the person who develops a FOF and should be considered when designing interventions.

Keywords

accidental falls injurious falls community-residing aged older adults

Background

Hip fracture rates are showing some reduction worldwide, yet hospitalization rates for falls have significantly increased over the past decade (Australian Institute of Health and Welfare, 2012). In the United States in 2000, the total direct medical costs of all fall injuries for people 65 and older exceeded $US19 billion (Stevens, Corso, Finkertein, & Miller, 2006) and, in Australia, have been reported as consuming 5% of health budgets (Watson, Clapperton, & Mitchell, 2010). Less known are the health costs from fear of falling (FOF), which, like injurious falls, can severely affect a person’s efficacy beliefs and lead to activity restriction, functional decline, and decreased quality of life (Scheffer, Schuurmans, van Dijk, van der Hooft, & de Rooij, 2008). Crucial to best investment and a broader reach of preventive interventions is that those at risk of an injurious fall or of FOF can access services and that interventions tackle the most modifiable risk factors from a population perspective.

Population studies of risk factors for injurious falls report a range of factors: cognitive impairment, multiple chronic conditions, balance and gait impairment, low body mass index (BMI; Tinetti, Doucette, Claus, & Marottoli, 1995), increasing age, low or very high activity levels, stroke, respiratory disorder (O’Loughlin, Robitaille, Boivin, & Suissa, 1993), past fractures, vision problems (Tromp, Smit, Deeg, Bouter, & Lips, 1998), urge incontinence (Brown et al., 2000), long sleep and daily naps (Stone et al., 2009), and differing results for gender being no difference (O’Loughlin et al., 1993) to being female of higher risk (Tromp et al., 1998). These studies have had follow-up periods ranging from 12 to 36 months. One study explored injurious falls over 8 years showing a decline in cognitive performance as an independent predictor (Anstey, von Sanden, & Luszcz, 2006).

Seldom are longitudinal studies designed to report incident falls. A large study (Ensrud et al., 2007) of women followed over 9 years examined the impact of frailty on incident falls and fractures. Although frailty was independently related to a risk of hip fracture, these results suggest that it is only weakly related to other types of fracture. Other studies reporting incident falls do not distinguish between non-injurious or injurious falls (Friedman, Munoz, West, Rubin, & Fried, 2002; Srikanth et al., 2009). There is a predominance of research on intrinsic factors; therefore, this study will broaden the scope to include a range of psychological and lifestyle factors.

FOF has been described as a general concept that captures low confidence in avoiding falls or being afraid of falling. The emerging research around predictors of FOF is based primarily on associations from cross-sectional studies, and few use population samples, thus resulting in limited knowledge about its temporal relationship to risk factors (Scheffer et al., 2008). One study that examined both incident falls and FOF concluded that the predictor factors were similar, though few factors were included and these were limited to intrinsic (Friedman et al., 2002). There is, as for injurious falls, limited understanding of the role of psychosocial and lifestyle factors in developing FOF.

We aimed to determine during an 11-year follow-up period in a regionally representative sample of older Australians whether a range of socio-demographic, health status, physical and functional statuses, and psychosocial and lifestyle factors predict having a fall requiring medical attention or of acquiring a FOF. We also examined whether having such a fall predicted FOF or vice versa.

Method

We used data from the Melbourne Longitudinal Studies on Healthy Ageing Program (MELSHA; Kendig, Helme, Teshuva, Osborne, & Flicker, 1996). The methodology was a prospective cohort study of 1,000 people conducted over 11 years from 1994 to 2005. The background, methodology, and instruments used are reported previously (Browning & Kendig, 2010).

Data Collection

The sampling frame was developed from data supplied by the Australian Electoral Commission using a clustered sampling strategy, excluding people who were living in non-private accommodation, who could not speak conversational English (11.3%) and who could not be interviewed at home for health reasons (3%). The response rate was 70.3% (1,000/1,422), being the number of eligible persons with complete data divided by the total number of eligible sampled persons. The baseline survey was a representative sample of 1,000 people aged 65 years and living in community dwelling at the time of the baseline interviews. Baseline data collection was conducted by 13 trained interviewers to standardize the interviewing procedure with 95% of baseline interviews completed from May to November 1994. Socio-medical data were gathered in face-to-face baseline interviews, and outcomes were identified in biennial follow-ups with respondents, informants, and death registries over 11 years. For participants who could not be re-contacted directly, the tracing procedures relied primarily on next of kin or other individuals volunteered by participants as key contacts. Death records were checked annually. Interview data were collected on 98% of the survivors over the study period.

Incident Falls Requiring Medical Attention and Incident-Acquired FOF

Participants were asked in biennial face-to-face interviews whether they had fallen in the last year, and if so, whether they received any medical treatment from injuries from these falls. They were also asked whether they were afraid of falling using a single global question with a 4-point forced choice response scale of not at all afraid, somewhat afraid, fairly afraid, or very afraid from the Frailty and Injuries: Cooperative Studies of Intervention Techniques (FICSIT) trials (Ory et al., 1993). Similar scales have been shown to have established content validity and moderate concurrent validity with self-efficacy scales (Yardley et al., 2005).

Predictor Variables

Table 1 summarizes the variables within the five block domains: socio-demographic, health status, physical and functional statuses, and psychosocial and lifestyle factors. Baseline data for BMI values were derived from measured height (to the nearest 0.5 cm) and weight (to the nearest kilogram) values. We classified weight according to the International Standards for BMI with a healthy weight 18.5 to 24.9, underweight <18.5, overweight 25 to 29.9, and obesity 30+.

Table 1.

Baseline Characteristics Listed Within the Five Domains (N = 1,000).

Baseline factor	M (range) or %
Socio-demographic
Age in years	73.4 (65-94)
Male gender	46.7%
Living alone	33.5%
Pre-tax weekly income^a	4.9 (1-11)
Ethnicity
Australian born	74.3%
Other English-speaking country	10.3%
European	11.9%
Asian/Other^b	3.5%
Health status
Fall requiring medical treatment in last year	9.6%
Self-rated health (highest poor health)	1.5 (0-4)
Self-rated health compared with 5 years ago (highest worse)	2.16 (0-4)
Hospital admission in previous year	21.8%
Tally of 10 fall-risk-related medical conditions^c	1.4 (0-6)
Tally of 23 non-fall-risk-related medical conditions	2.43 (0-10)
Degree of cognitive impairment (0-9, highest impaired)^d	0.87 (0-5)
Physical status
Timed Get Up and Go (s)^e	12.28 (4-40)
Total physical activity (highest active)^f	8.64 (0-25)
Physical functioning composite^g (0-6, highest higher functioning)	3.61 (0-5.83)
BMI^h
Underweight	5.7%
Acceptable	27.3%
Overweight	49.9%
Obese	17.1%
Sleeping problems (range, highest problems)	6.36 (3-12)
Problems with vision	1.06 (0-2.7)
Urinary incontinence problems	24.3%
Pain last 12 monthsⁱ (range, highest daily)	2.6 (1-5)
Difficulty getting around	31.4%
IADL dependence (dependent)	17.5%
I/ADL help received (0-6, highest more help)	0.30 (0-5)
Psychosocial
Afraid of falling^j	14.5%
Positive affect (highest positive)	15.10 (4-20)
Negative affect (highest negative)	7.93 (0-20)
Depression scale score^k (highest depressed)	1.38 (0-11)
Lifestyle
Reduced regular social activity in past 5 years	32.4%
Social activity past 2 weeks (0-9, highest socially active)	2.82 (0-8)
Easy walking neighborhood (highest more difficulty)	0.90 (0-4)
Main means of transport
Car	75.2%
Public transport	18.6%
Walking	6.2%
Total nutritional score^l (highest good nutrition)	14.02 (0-19)
Healthy appetite/eating (highest good)	1.43 (0.25-2)

Note. BMI = body mass index; IADL = Instrumental Activities of Daily Living.

One to 11 weekly income categories ranging from $AU0 to $AU926 or more.

Country of birth: Asia, 1.7%; Africa, 1.3%; Middle East, 0.1%; Other, 0.2%.

Ten fall-risk medical conditions: Stroke, osteoporosis, gout, other arthritis, problems with feet or legs, cataracts, glaucoma, emphysema, Parkinson’s disease.

Derived from the Organic Brain Syndrome scale.

Timed Get Up and Go test.

Self-report of frequency of light, moderate, or energetic physical activity done in past 2 weeks.

Composite physical functioning score derived from above.

BMI score: Underweight < 18.5, acceptable 18.5 to 24.9, overweight 25 to 29.9, obese (30+).

FICSIT global measure of fear of falling; not at all afraid (No), somewhat afraid, fairly afraid, or very afraid (Yes).

Psychiatric Assessment Depression subscale.

Daily pain, once or twice a week, once or twice per month, a few times a year, never.

Weekly consumption of fruits, vegetables, salads, and milk products use to create composite score.

Functional mobility (gait and balance) was assessed using the timed “Get Up and Go” test (Podsiadlo & Richardson, 1991). Classification of medical conditions was by self-report as ever having been medically diagnosed or by specific prescription medications currently being used, as determined at the baseline interview. Depressive symptoms in the past 2 weeks were assessed using the Psychogeriatric Assessment Scales (PAS), depression scale. Independence with activities such as gardening and home maintenance, shopping for groceries, preparing own meals, doing own housework, taking own medicine, and getting to places within walking distance was assessed using the Instrumental Activities of Daily Living (IADL) sections of the Older Americans Resources and Services (OARS) Multidimensional Functional Assessment Questionnaire, and the Multi-Level Functional Assessment Instrument. Cognitive impairment was assessed using components of the Organic Brain Syndrome (OBS) scale (Gustafson, Lindgren, & Westling, 1995). Data collected by questionnaire are summarized as footnotes in Table 1. Background and cohort profile of the Melbourne Longitudinal Study on Healthy Aging is summarized in Browning and Kendig (2010).

Data Analysis

Two analyses were conducted, one excluding people at baseline who reported having a injurious fall in the previous year and the other removing those who reported a FOF at baseline. We then examined predictors of incident injurious falls and FOF over the 11 years of follow-up.

For constructed scales, missing values on component items were pro-rated by the total on the rest of the items. Missing values were imputed for each of these constructs by using (mean-substitution) multiple linear regression of the construct of interest on other relevant baseline predictors. Data were analyzed using a Cox proportional hazards (Cox regression) method with a hierarchical model. In the first step, all variables were entered as single predictors with a significance level of p < .01 as criteria for further inclusion. The next run was a five block model of the socio-demographic, health status, physical status, and psychosocial and lifestyle factors meeting the inclusion criteria with a p < .01 criterion for inclusion in the final model. In the final model, we accepted factors with a significance at p < .05. Proportions of hazards assumptions were based on Cox and Snell:

R^{2} = 1 - e^{- G^{2} / n},

where G² is the change in chi square associated with the model (Tabachnick & Fidell, 2002).

Results

Table 1 outlines the baseline characteristics and variables within each of the five blocks. The incidence of experiencing an injurious fall for the first time in follow-up was 22.1% (n = 200; see summary in Figure 1a), and of these, 20% (n = 40) had two or more injurious falls. There was a mean survival time to first report of an injurious fall of 5.1 years (range = 0-11). The incidence of experiencing a FOF for the first time in follow-up was 13.7% (n = 117; see Figure 1b), and of these, 33% (n = 28) reverted back to a no FOF condition during follow-up. There was a mean survival time to first report of FOF of 5.6 years (range = 0-11.0). At the end of the 11-year period, 513 people had either entered a nursing home or died. Table 2 summarizes the block models of predictors for the injurious falls analysis and Table 3 for the FOF analysis.

Figure 1.

Person’s with incident events over 11 years follow-up (N = 1,000).

Table 2.

Baseline Predictors of Incident Falls Requiring Medical Treatment (Injurious Falls) Over 11 Years (n = 904).

	Single predictor (p = .10)		Block model (p = .10)		Final model (p = .05)		95% CI
Variable^a	HR	P	HR	p	HR	p	Lower	Upper
Socio-demographic
Age in years	1.06	.000	1.06	.000	1.06	.000	1.03	1.087
Male gender indicator	0.62	.001	0.66	.005
Living alone	1.40	.024
Health status
Tally of 10 fall-risk-related medical conditions	1.15	.022
Tally of 23 non-fall-risk-related medical conditions	1.12	.007	1.12	.007
Physical and functional status
Timed walk (s)	1.03	.001	1.03	.000	1.02	.010	1.01	1.043
BMI: Underweight-acceptable	0.49	.018	0.45	.004	0.53	.023	0.30	0.917
Pain last 12 months	1.08	.072
Psychosocial
Afraid of falling	1.61	.012
Positive affect	0.93	.011
Negative affect	1.06	.014
Depression scale score	1.14	.000	1.14	.000	1.12	.001	1.05	1.202
Lifestyle
Main transport: Public car	1.47	.034	1.38	.073
Healthy appetite/eating	0.59	.013	0.61	.019

Note. Final block model: G² = 42.75; Cox and Snell R² = 12.8%. CI = confidence interval; HR = hazard ratio; BMI = body mass index.

Variables in Table 1 that are not in Table 2 failed to meet criterion in single predictor analysis, which included the variable fear of falling.

Table 3.

Baseline Predictors of Acquiring Fear of Falling Over 11 Years (n = 855).

	Single predictor (p = .10)		Block model (p = .10)		Final model (p = .05)		95% CI
Variable^a	HR	p	HR	p	HR	p	Lower	Upper
Socio-demographic
Age in years	1.09	.000	1.09	.000	1.09	.000	1.05	1.13
Male gender indicator	0.34	.000	0.35	.000	0.38	.000	0.25	0.57
Living alone	1.44	.061
Pre-tax weekly income	0.86	.010
Ethnicity Asian/Other cf. Australian	2.81	.009	4.43	.000	4.08	.001	1.83	9.10
Health status
Self-rated health	1.36	.001
Rated health compared with 5 years ago	1.32	.022
Tally of 10 fall-risk medical conditions	1.39	.000	1.35	.000
Tally of 23 non-fall-risk medical conditions	1.11	.043
Cognitive impairment (0-9)	1.54	.000	1.50	.000	1.45	.000	1.22	1.73
Physical and functional statuses
Timed walk (s)	1.04	.000
Physical functioning score (0-6)	0.59	.000	0.61	.000	0.64	.000	0.53	0.77
BMI: Underweight cf. acceptable	0.20	.000	0.25	.000	0.43	.004	0.24	0.76
Sleeping problems	1.17	.000	1.12	.009
Problems with vision	1.61	.026
Urinary incontinence problems	1.69	.021
Difficulty getting around	1.67	.011
IADL dependence (0-1)	2.64	.000
I/ADL help received (0-6)	1.50	.000
Psychosocial
Positive affect	0.90	.003
Negative affect	1.09	.007
Depression scale score	1.24	.000	1.24	.000
Lifestyle
Reduced regular social activity in past 5 years	2.08	.000	1.77	.003	1.81	.002	1.25	2.62
Main transport: Public car	2.10	.000	2.10	.000
Total nutritional score	0.90	.001	0.93	.011
Healthy appetite/eating	0.31	.000	0.43	.003

Note. Final block model: G² = 120.76, Cox and Snell R² = 12.5%. CI = confidence interval; HR = hazard ratio; BMI = body mass index; IADL = Instrumental Activities of Daily Living; cf = other than.

Variables in Table 1 that are not in Table 3 failed to meet criterion in single predictor analysis, which included the variable having an injurious fall.

Predictors of Falls Requiring Medical Treatment

Predictors of falls requiring medical treatment, n = 904, events (reporting one or more injurious falls) = 200, were increasing age; a measure of frailty, which was the timed Get Up and Go in seconds; and being in a state of depression. Being underweight compared with an acceptable weight measured by a BMI was protective of having an injurious fall (see Table 2).

Variables that were significant (p < .10) in the first block model but not in the final model were male gender (hazard ratio [HR] = 0.66), composite of non-falls-related medical conditions (HR = 1.12), using public transport instead of a car (HR = 1.38), and healthy appetite and eating (HR = 0.61).

FOF

Predictors of FOF, n = 855, events (reporting a FOF) = 117 people, were increasing age; being from a non-English, Australian, or European background; having a degree of cognitive impairment; and reporting reduced social activity during the past 5 years (see Table 3). Protective factors were high physical functional capacity, being male, and having a low BMI compared with those of acceptable weight.

Variables that were significant (p < .10) in the first-block model but not in the final model were a composite of fall-related medical conditions (HR = 1.35), sleeping problems (HR = 1.12), depression (HR = 1.24), using public transport compared with a car (HR = 2.10), nutritional score (HR = 0.93), and healthy appetite/eating (HR = 0.43).

Discussion

We examined baseline characteristics and then followed a cohort of community-dwelling people to determine who would seek medical assistance for an injurious fall and who would develop a FOF. Our results provide novel evidence that having an injurious fall does not predict acquiring a FOF nor did FOF predict a future injurious fall. Furthermore, the profiles of the person who will have an injurious fall differ to the profile of the person who develops a FOF. Our findings are strong indicators that injurious falls and FOF are different phenomena and supports different approaches are needed when designing interventions.

There is much debate about the causal relationship between FOF and injurious falls and a continuing assumption that they are always inter-connected. Earlier theories posit that falling leads to a FOF, which leads to activity restriction, which then leads to deterioration of physical fitness and capacity, social isolation, depression, and then an increased risk of falling. This traditional and “one scenario” belief still dominates our understanding. However, our study, and others, has demonstrated that FOF is evident without reported falls or injury from falling. Our study reinforces this demonstrating over 11 years that one does not predict the other. Delbaere, Crombez, Van Den Noorgate, Willems, and Cambier (2006) were the first to raise the notion that people with high fear or no concern about falling can display either inappropriate or appropriate responses to their actual physiological risk for falling. More recently, there has been argument that activity restriction is not necessarily directly on the causal pathway between FOF and falling and that balance performance has a greater mediation role (Hadjistavropoulos, Delbaere, & Fitzgerald, 2011). It appears that functional capacity in terms of gait and balance maybe a stronger predictor of falls than FOF. Furthermore, people’s perception of activity restriction, a known consequence of both falls and FOF, is in contrast to evidence around what people actually do. It seems that engagement in the necessary tasks of daily life activity may remain whether or not they have a FOF (Hornyak et al., 2013) or experience falls (Chan et al., 2007). But this is more complex as these studies do not examine activity engagement from a perspective of meaningful activity or social interaction. The limitations previously from longitudinal studies have concentrated on intrinsic factors, and function is most often measured from a limited self-care aspect.

In the following, we first discuss the predictors of onset of injurious falls and then those who acquired a FOF. Increasing age as a predictor of injurious falls was not an unexpected finding as it is consistently shown in risk factor studies to be associated with falling. Much has been studied about gait speed and falls, and our findings, as have others, support the use of gait speed as a useful screen for falls. Gait stability can be affected by decreased strength, and range of motion at a variety of joints; neurological factors, executive function, and confidence. A meta-analysis of fall-risk factor studies showed gait problems to be within the highest group of risk with over a twofold chance of falling (OR = 2.02 [1.30-2.93]; Deandrea et al., 2010), and gait impairment has also been shown to be an independent risk factor for injurious falls (Koski, Luukinen, Laippala, & Kivela, 1998).

An important finding was that depression is an independent predictor of new onset injurious falls requiring medical attention, thus extending the findings of Anstey, Burns, von Sanden, and Luszcz (2008) from a population perspective, which demonstrated that depression, moral, and control were independent predictors of falling over an 8-year follow-up. Explanatory risks factors may be due to both cognitive and physical impairments. These may include slower walking speed, poorer balance, reduced strength related to inactivity to influences of impaired executive functioning on attention and judgment (Ble et al., 2005; Liu-Ambrose, Nagamatsu, Hsu, & Bolandzadeh, 2013; Springer et al., 2006). These findings support screening of depression in those who fall, the inclusion of psychological well-being and motivation as a focus within fall prevention interventions, and the link between exercise and mood as particularly important.

The finding that being underweight measured by low BMI was protective of having an injurious fall is contrary to other findings of BMI as a risk factor for fracture (De Laet et al., 2005) and mortality (Atlantis, Browning, & Kendig, 2010). A longitudinal study (Himes & Reynolds, 2012) found that falls increase with obesity but that injurious falls show a different pattern. This study provides support for our unexpected finding and, also contrary to their hypothesis, found that low BMI was not predictive of injurious falls. Being under- or overweight is likely associated with inactivity and subsequent less exposure to falling. Another explanation may be in the use of BMI, which has been criticized as it does not reflect muscle quality and that sarcopenia, which is the loss of muscle mass associated with functional decline measured by bioimpedance analysis, is not related to BMI in older people (Janssen, Heymsfield, & Ross, 2002).

We found that predictors of acquiring a FOF include increasing age, poorer physical functioning, gender, decreasing social engagement, and having some degree of cognitive impairment. Another significant factor, though a wide confidence interval reflecting small numbers, was being born in Asia or Africa. Being underweight compared with an acceptable weight was, as in our results for fallers, protective of developing a FOF. Our incident rate of developing FOF (13.7%) was comparable with other population-based studies such as Friedman et al.’s (2002) 16% but less than the rate of 27% by Murphy, Dubin, and Gill (2003) in their women-only population sample, which makes sense as women are more at risk of FOF. Women’s propensity to have a higher rate of FOF may be due more to physical functional decline than psychological issues (Martin, Hart, Spector, Doyle, & Harari, 2005). We did not find depression to be a predictor for FOF although depression does appear to be a consequence of long-standing FOF (Austin, Devine, Dick, Prince, & Bruce, 2007).

Studies have shown that restriction of activity is a consequence of FOF (Scheffer et al., 2008), and there has been some discussion with regard to which comes first (Hadjistavropoulos et al., 2011). Ours is the first study to show that a reduction in the person’s regular social activities can predict (HR = 1.8) a future onset of FOF in the years ahead. This is a new finding and points to the importance of tackling FOF within a person-centered framework to support engaging in social and meaningful activity. Murphy et al. (2003), although having a small sample and only women, showed that having a sedentary lifestyle and reporting a lack of emotional support were both predictive of FOF. They captured sedentary lifestyle in terms of lack of engagement in terms of exercise, sport, avoiding stair climbing, and limited outdoor walks. FOF can have severe consequences that could be avoidable. Enabling social and activity engagement and preventing social isolation may help prevent the onset of FOF.

Mild cognitive impairment (MCI) may give rise to reduced coping skills and impairment of insight or capacity to realistically appraise fall risk. Evidence toward the connection between MCI and social isolation is emerging showing a reduction of activities that involve social connectedness and community activity (Kaye, Matteck, Hayes, Austin, & Dodge, 2012). The accelerating increase in the prevalence of dementia, with costs attributed to 1% or more of Gross Domestic Product (GDP) in developed countries, provides a crucial impetus to address vulnerable people with MCI in the community (Alzheimer’s Disease International, 2011).

Our finding of a higher incidence of onset FOF among Asian, African, and other non-European migrants requires further research. Studies in Asian countries are emerging and have found the rate of FOF in some countries to be substantially higher than falls (Lim et al., 2011). Cultural influences on health beliefs and on exercise and physical activity for migrant groups may well affect the prevalence of FOF in these groups and need further investigation. These differences have important implications for focused interventions, particularly important in countries with high immigrant groups. In countries like Australia, Asian immigrants will become a very significant part of the younger-old, with implications for health services provision and access (Karmel, Lloyd, & Hales, 2007).

We reported being underweight as protective of FOF, while Austin et al. (2007) reported obesity as predictive of FOF compared with an acceptable weight. These findings may be related to levels of physical activity but, as for our fall findings, caution should be applied due to the limitation of a small sub-sample of those with low BMI, which is reflected in the wide confidence interval.

Gait and other intrinsic factors were not predictors of FOF but are identified as risks for FOF in earlier and mostly cross-sectional studies. The assertion is that they are associated with those people who fall and also have a FOF (Scheffer et al., 2008). It has been argued (Hadjistavropoulos et al., 2011) that if FOF and falls are not independent predictors for each other, which we support in our findings, then for those who experience both, which is about half, it is likely that the relationship between them is mediated by intrinsic factors such as gait, balance, and medication side effects. This is another question that deserves exploration.

One population study (Friedman et al., 2002) examined both falls and FOF, although falls were not limited to those requiring medical attention, and recall data were collected at one time point only. In contrast to Friedman et al., we found that FOF and falling had different predictors. Unlike their study, general health status was not a predictor and did not hold over our 11-year follow-up period. The major difference was that we broadened the understanding by including a range of psychosocial and lifestyle factors. We also did not find injurious falls with medical follow-up to be a predictor of FOF or the reverse. In Friedman’s study, falls captured all falls, whereas in this analysis, we focused on fallers at risk, that is, having an injurious fall. Furthermore, their FOF analysis also delineated those with activity restriction or not and it was those with both FOF and activity restriction at baseline that predicted incident falls. This is further evidence that suggests the importance of intervening in FOF through enabling engagement in activity.

One important risk factor for falls not included in this study was multiple or psychotropic medication use, although conflicting results have been reported for FOF (Denkinger, Lukas, Nikolaus, & Hauer, 2014). Another limitation is the long period of follow-up that can contribute to selection effects and the sample increasingly being “survivors.” Recall of falls introduces a bias as people are more likely to recall a fall linked to an event rather than a specific time frame. The advantage for this study is that falls that result in injury and in medical follow-up are much more likely to be recalled (Ganz, Higashi, & Rubenstein, 2005). We chose an injurious fall requiring medical attention as we believed this set a threshold high enough to be of health significance. Bi-annual reporting of injurious falls with a 1-year recall period does increase the risk of bias and would result in an underestimation of falls. Strengths include the population representative sample, trained assessors in face-to-face interviews, and the use of validated psychometric scales such as for depression and cognitive impairment and an objective measure for gait speed.

In summary, we have provided insights into how the profiles of people who have a fall requiring medical attention and those who develop a FOF differ and who could benefit from intervention. These findings delineate programs that offer strategies to reduce FOF and increase activity levels such as Matter of Balance (Howland et al., 1993) and the more recent multicomponent cognitive behavioral group intervention by Zijlstra et al. (2011) in contrast to group programs aimed to reduce falls such as Stepping On (Clemson et al., 2004). It seems there is a place for both. We challenged post-event avoidance models (Hadjistavropoulos et al., 2011) by showing reduced social activity is a precursor of FOF and suggest other models may need exploration. Further work is needed to investigate interventions for people with depression, cognitive impairment, and possibly from specific cultural sub-groups. Finally, interventions to reduce falling require more understanding of the significance of gait, balance, and psychosocial approaches; those that target FOF need to consider lifestyle factors. The findings have broader implications for active aging to promote engagement in life activities as well as reablement and restoration programs in the community that are crucial to improving the lives of older at-risk people as they age (Lewin, Vandermeulen, & Coster, 2006).

Footnotes

Acknowledgements

The authors thank the study participants for their long-term commitment to the project.

Authors’ Note

Funding bodies had no role in the design and conduct of this study.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The Melbourne Longitudinal Studies on Healthy Ageing (MELSHA) program has been funded by a large number of grants and supporting agencies. They include the Victorian Health Promotion Foundation, the National Health and Medical Research Council (NHMRC), and the Australian Research Council. Professor Clemson is supported by an NHMRC Career Development Fellowship.

References

Alzheimer’s Disease International. (2011). World Alzheimer’s report 2010: The global economic impact of dementia. Retrieved from http://www.alz.co.uk/research/files/WorldAlzheimerReport2010.pdf

Anstey

K. J.

Burns

von Sanden

Luszcz

M. A.

(2008). Psychological well-being is an independent predictor of falling in an 8-year follow-up of older adults. Journals of Gerontology, Series B: Psychological Sciences & Social Sciences, 63, P249-P257.

Anstey

K. J.

von Sanden

Luszcz

M. A.

(2006). An 8-year prospective study of the relationship between cognitive performance and falling in very old adults. Journal of the American Geriatrics Society, 54, 1169-1176.

Atlantis

Browning

Kendig

(2010). Body mass index and unintentional weight change associated with all-cause mortality in older Australians: The Melbourne Longitudinal Studies on Healthy Ageing (MELSHA). Age and Ageing, 39, 643-646.

Austin

Devine

Dick

Prince

Bruce

(2007). Fear of falling in older women: A longitudinal study of incidence, persistence, and predictors. Journal of the American Geriatrics Society, 55, 1598-1603.

Australian Institute of Health and Welfare. (2012). Australia’s health 2012. Canberra: Author.

Ble

Volpato

Zuliani

Guralnik

J. M.

Bandinelli

Lauretani

. . . Ferrucci

(2005). Executive function correlates with walking speed in older persons: The InCHIANTI study. Journal of the American Geriatrics Society, 53, 410-415.

Brown

J. S.

Vittinghoff

Wyman

J. F.

Stone

K. L.

Nevitt

M. C.

Ensrud

K. E.

Grady

(2000). Urinary incontinence: Does it increase risk for falls and fractures? Study of Osteoporotic Fractures Research Group. Journal of the American Geriatrics Society, 48, 721-725.

Browning

C. J.

Kendig

(2010). Cohort profile: The Melbourne longitudinal studies on healthy ageing program (Research support, Non-U.S. Gov’t). International Journal of Epidemiology, 39(5), e1-e7.

10.

Chan

B. K. S.

Marshall

L. M.

Winters

K. M.

Faulkner

K. A.

Schwartz

A. V.

Orwoll

E. S.

(2007). Incident fall risk and physical activity and physical performance among older men: The Osteoporotic Fractures in Men Study. American Journal of Epidemiology, 165, 696-703.

11.

Clemson

Cumming

R. G.

Kendig

Swann

Heard

Taylor

(2004). The effectiveness of a community-based program for reducing the incidence of falls among the elderly: A randomized trial. Journal of American Geriatrics Society, 52, 1487-1494.

12.

Deandrea

Lucenteforte

Bravi

Foschi

La Vecchia

Negri

(2010). Risk factors for falls in community-dwelling older people: A systematic review and meta-analysis. Epidemiology, 21, 658-668.

13.

De Laet

Kanis

J. A.

Oden

Johanson

Johnell

Delmas

. . . Tenenhouse

(2005). Body mass index as a predictor of fracture risk: A meta-analysis. Osteoporosis International, 16, 1330-1338.

14.

Delbaere

Crombez

Van Den Noorgate

Willems

Cambier

(2006). The risk of being fearful or fearless of falls in older people: An empirical validation. Disability and Rehabilitation, 28, 751-756.

15.

Denkinger

Lukas

Nikolaus

Hauer

(2014). Factors associated with fear of falling and associated activity restriction in community dwelling older adults: A systematic review. The American Journal of Geriatric Psychiatry. Advance online publication. doi:10.1016/j.jagp.2014.03.002

16.

Ensrud

K. E.

Ewing

S. K.

Taylor

B. C.

Fink

H. A.

Stone

K. L.

Cauley

J. A.

. . . Cawthon

P. M.

(2007). Frailty and risk of falls, fracture, and mortality in older women: The study of osteoporotic fractures. Journals of Gerontology, Series A: Biological Sciences & Medical Sciences, 62, 744-751.

17.

Friedman

S. M.

Munoz

West

S. K.

Rubin

G. S.

Fried

L. P.

(2002). Falls and fear of falling: Which comes first? A longitudinal prediction model suggests strategies for primary and secondary prevention. Journal of the American Geriatrics Society, 50, 1329-1335.

18.

Ganz

Higashi

Rubenstein

(2005). Monitoring falls in cohort studies of community-dwelling older people: Effect of the recall interval. Journal of the American Geriatrics Society, 53, 2190-2194.

19.

Gustafson

Lindgren

Westling

(1995). The OBS scale: A factor analysis approach to evaluation of confused states and other organic brain syndromes. Lund, Sweden: Department of Psychogeriatrics, Lund University.

20.

Hadjistavropoulos

Delbaere

Fitzgerald

T. D.

(2011). Reconceptualizing the role of fear of falling and balance confidence in fall risk. Journal of Aging and Health, 23, 3-23.

21.

Himes

C. L.

Reynolds

S. L.

(2012). Effect of obesity on falls, injury, and disability. Journal of the American Geriatrics Society, 60, 124-129.

22.

Hornyak

Brach

J. S.

Wert

D. M.

Hile

Studenski

Vanswearingen

J. M.

(2013). What is the relation between fear of falling and physical activity in older adults? Archives of Physical Medicine and Rehabilitation, 94, 2529-2534.

23.

Howland

Peterson

E. W.

Levin

W. C.

Fried

Pordon

Bak

(1993). Fear of falling among the community-dwelling elderly. Journal of Aging and Health, 5, 229-243.

24.

Janssen

Heymsfield

S. B.

Ross

(2002). Low relative skeletal muscle mass (sarcopenia) in older persons is associated with functional impairment and physical disability. Journal of the American Geriatrics Society, 50, 889-896.

25.

Karmel

Lloyd

Hales

(2007). Older Australians at a glance (Bulletin No. 53). Canberra: Australian Institute of Health and Welfare.

26.

Kaye

Matteck

Hayes

Austin

Dodge

(2012, July 17). Community life withdrawal with MCI progression. Paper presented at the Alzheimer’s Association International Conference 2012, Vancouver, British Columbia, Canada.

27.

Kendig

Helme

Teshuva

Osborne

Flicker

(1996). Health status of older people project: Preliminary findings of a survey of the health and lifestyles of older Australians. Melbourne, Australia: Victorian Health Promotion Foundation.

28.

Koski

Luukinen

Laippala

Kivela

S. L.

(1998). Risk factors for major injurious falls among the home-dwelling elderly by functional abilities: A prospective population-based study (Research support, Non-U.S. Gov’t). Gerontology, 44, 232-238.

29.

Lewin

Vandermeulen

Coster

(2006). Programs to promote independence at home: How effective are they? Generations Review, 16(3/4), 24-26.

30.

Lim

J.-Y.

Jang

S.-N.

Park

W.-B.

M. K.

Kang

E. K.

Paik

N.-J.

(2011). Association between exercise and fear of falling in community-dwelling elderly Koreans: Results of a cross-sectional public opinion survey. Archives of Physical Medicine and Rehabilitation, 92, 954-959.

31.

Liu-Ambrose

Nagamatsu

L. S.

Hsu

L. C.

Bolandzadeh

(2013). Emerging concept: “Central benefit model” of exercise in falls prevention. British Journal of Sports Medicine, 47(2), 115-117.

32.

Martin

F. C.

Hart

Spector

Doyle

D. V.

Harari

(2005). Fear of falling limiting activity in young-old women is associated with reduced functional mobility rather than psychological factors. Age and Ageing, 34, 281-287.

33.

Murphy

S. L.

Dubin

J. A.

Gill

T. M.

(2003). The development of fear of falling among community-living older women: Predisposing factors and subsequent fall events. Journals of Gerontology, Series A: Biological Sciences & Medical Sciences, 58, M943-M947.

34.

O’Loughlin

J. L.

Robitaille

Boivin

J. F.

Suissa

(1993). Incidence of and risk factors for falls and injurious falls among the community-dwelling elderly. American Journal of Epidemiology, 137, 342-354.

35.

Ory

Schechtman

Miller

Hadley

Fiatarone

Province

. . .Kaplan

(1993). The fear of falling scale from the frailty and injuries: Cooperative studies of intervention techniques (FICSIT) trials. Journal of the American Geriatrics Society, 41, 283-296.

36.

Podsiadlo

Richardson

(1991). The timed “up & go”: A test of basic functional mobility for frail elderly persons. Journal of the American Geriatrics Society, 39, 142-148.

37.

Scheffer

A. C.

Schuurmans

M. J.

van Dijk

van der Hooft

de Rooij

S. E.

(2008). Fear of falling: Measurement strategy, prevalence, risk factors and consequences among older persons (Review). Age and Ageing, 37, 19-24.

38.

Springer

Giladi

Peretz

Yogev

Simon

E. S.

Hausdorff

J. M.

(2006). Dual-tasking effects on gait variability: The role of aging, falls, and executive function. Movement Disorders, 21, 950-957.

39.

Srikanth

Beare

Blizzard

Phan

Stapleton

Chen

Reutens

(2009). Cerebral white matter lesions, gait, and the risk of incident falls: A prospective population-based study (Research support, Non-U.S. Gov’t). Stroke, 40, 175-180.

40.

Stevens

J. A.

Corso

P. S.

Finkertein

E. A.

Miller

T. R.

(2006). The costs of fatal and non-fatal falls among older adults. Injury Prevention, 12, 290-295.

41.

Stone

K. L.

Ewing

S. K.

Ancoli-Israel

Ensrud

K. E.

Redline

Bauer

D. C.

. . .Cummings

S. R.

(2009). Self-reported sleep and nap habits and risk of mortality in a large cohort of older women. Journal of the American Geriatrics Society, 57, 604-611.

42.

Tabachnick

B. G.

Fidell

L. S.

(2002). Using multivariate statistics (4th ed.). Boston, MA: Allyn & Bacon.

43.

Tinetti

M. E.

Doucette

Claus

Marottoli

(1995). Risk factors for serious injury during falls by older persons in the community. Journal of the American Geriatrics Society, 43, 1214-1221.

44.

Tromp

A. M.

Smit

J. H.

Deeg

D. J.

Bouter

L. M.

Lips

(1998). Predictors for falls and fractures in the Longitudinal Aging Study Amsterdam. Journal of Bone and Mineral Research, 13, 1932-1939.

45.

Watson

Clapperton

Mitchell

R. T.

(2010). The incidence and cost of falls injury among older people in New South Wales 2006/07. Sydney, Australia: NSW Department of Health. Available from www.health.nsw.gov.au

46.

Yardley

Beyer

Hauer

Kempen

Piot-Ziegler

Todd

(2005). Development and initial validation of the Falls Efficacy Scale-International (FES-I). Age and Ageing, 34, 614-619.

47.

Zijlstra

G. A. R.

van Haastregt

J. C. M.

van Eijk

J. T. M.

de Witte

L. P.

Ambergen

Kempen

G. I. J. M.

(2011). Mediating effects of psychosocial factors on concerns about falling and daily activity in a multicomponent cognitive behavioral group intervention. Aging & Mental Health, 15, 68-77.